Simulation-based CI for slope

Suy luận cho hồi quy tuyến tính trong R

Jo Hardin

Professor, Pomona College

Permutation vs. bootstrap variability

Slopes from permuted data

Slopes from bootstrapped data

Permutation vs. bootstrap code

Permutation:

twins %>% 
  specify(Foster ~ Biological) %>% 
  hypothesize(null = "independence") %>% 
  generate(reps = 100, type = "permute") %>% 
  calculate(stat = "slope")

Bootstrap:

twins %>% 
  specify(Foster ~ Biological) %>%  
  generate(reps = 100, type = "bootstrap") %>% 
  calculate(stat = "slope")

Sampling distribution: randomization vs. bootstrap

Slopes from permuted data

Slopes from bootstrapped data

Let's practice!

Suy luận cho hồi quy tuyến tính trong R